Security News
Oracle Drags Its Feet in the JavaScript Trademark Dispute
Oracle seeks to dismiss fraud claims in the JavaScript trademark dispute, delaying the case and avoiding questions about its right to the name.
URIS
http://rubyforge.org/projects/codeforpeople/ http://www.codeforpeople.com/lib/ruby/
SYNOPSIS
open child process with handles on pid, stdin, stdout, and stderr: manage child processes and their io handles easily.
INSTALL
~> gem install open4
SAMPLES
harp: > cat sample/simple.rb
require "open4"
pid, stdin, stdout, stderr = Open4::popen4 "sh"
stdin.puts "echo 42.out"
stdin.puts "echo 42.err 1>&2"
stdin.close
ignored, status = Process::waitpid2 pid
puts "pid : #{ pid }"
puts "stdout : #{ stdout.read.strip }"
puts "stderr : #{ stderr.read.strip }"
puts "status : #{ status.inspect }"
puts "exitstatus : #{ status.exitstatus }"
harp: > ruby sample/simple.rb
pid : 17273
stdout : 42.out
stderr : 42.err
status : #<Process::Status: pid=17273,exited(0)>
exitstatus : 0
harp: > cat sample/block.rb
require 'open4'
status =
Open4::popen4("sh") do |pid, stdin, stdout, stderr|
stdin.puts "echo 42.out"
stdin.puts "echo 42.err 1>&2"
stdin.close
puts "pid : #{ pid }"
puts "stdout : #{ stdout.read.strip }"
puts "stderr : #{ stderr.read.strip }"
end
puts "status : #{ status.inspect }"
puts "exitstatus : #{ status.exitstatus }"
harp: > ruby sample/block.rb
pid : 17295
stdout : 42.out
stderr : 42.err
status : #<Process::Status: pid=17295,exited(0)>
exitstatus : 0
harp: > cat sample/exception.rb
require "open4"
Open4::popen4 "noexist"
harp: > ruby sample/exception.rb
/dmsp/reference/ruby-1.8.1//lib/ruby/site_ruby/open4.rb:100:in `popen4': No such file or directory - noexist (Errno::ENOENT)
from sample/exception.rb:3
harp: > cat sample/spawn.rb
require 'open4'
include Open4
cat = ' ruby -e" ARGF.each{|line| STDOUT << line} " '
stdout, stderr = '', ''
status = spawn cat, 'stdin' => '42', 'stdout' => stdout, 'stderr' => stderr
p status
p stdout
p stderr
stdout, stderr = '', ''
status = spawn cat, 0=>'42', 1=>stdout, 2=>stderr
p status
p stdout
p stderr
harp: > RUBYLIB=lib ruby sample/spawn.rb
0
"42"
""
0
"42"
""
harp: > cat sample/bg.rb
require 'yaml'
require 'open4'
include Open4
stdin = '42'
stdout = ''
stderr = ''
t = bg 'ruby -e"sleep 4; puts ARGF.read"', 0=>stdin, 1=>stdout, 2=>stderr
waiter = Thread.new{ y t.pid => t.exitstatus } # t.exitstatus is a blocking call!
while((status = t.status))
y "status" => status
sleep 1
end
waiter.join
y "stdout" => stdout
harp: > ruby sample/bg.rb
---
status: run
---
status: sleep
---
status: sleep
---
status: sleep
---
21357: 0
---
stdout: "42\n"
harp: > cat sample/stdin_timeout.rb
require 'open4'
producer = 'ruby -e" STDOUT.sync = true; loop{sleep(rand+rand) and puts 42} "'
consumer = 'ruby -e" STDOUT.sync = true; STDIN.each{|line| puts line} "'
open4(producer) do |pid, i, o, e|
open4.spawn consumer, :stdin=>o, :stdout=>STDOUT, :stdin_timeout => 1.4
end
harp: > ruby sample/stdin_timeout.rb
42
42
42
42
42
/dmsp/reference/ruby-1.8.1//lib/ruby/1.8/timeout.rb:42:in `relay': execution expired (Timeout::Error)
harp: > cat sample/pfork4.rb
require 'open4'
echo = lambda do
$stdout.write $stdin.read
raise 'finish implementing me'
end
org_message = "hello, world!"
got_message = nil
exception = nil
begin
Open4.pfork4(echo) do |cid, stdin, stdout, stderr|
stdin.write org_message
stdin.close
got_message = stdout.read
end
rescue RuntimeError => e
exception = e.to_s
end
puts "org_message: #{org_message}"
puts "got_message: #{got_message}"
puts "exception : #{exception}"
harp: > ruby sample/pfork4.rb
org_message: hello, world!
got_message: hello, world!
exception : finish implementing me
HISTORY 1.0.0 - added ability for spawn to take a proc (respond_to?(:call))
cmd = ' ruby -e" 42.times{ puts 0b101010 } " '
include Open4
spawn cmd, :stdout => lambda{|buf| puts buf}
0.9.5: - another patch from Corey Jewett, this time dealing with ruby's handling of chdir and threads. basically the 'cwd' keyword to open4 cannot work with multiple threads (aka background) because ruby cannot cause green threads to have an actuall different working dir. the moral is that the :cwd/'cwd' keyword to spawn will work with 0 or 1 threads in effect.
0.9.4: - patch to #background from Corey Jewett
0.9.3: - removed some debugging output accidentally left in 0.9.2. arggh!
0.9.2: - fixed a descriptor leak. thanks Andre Nathan.
0.9.1: - fixed warning with '-w' : @cid not initialized. thanks blaise tarr.
0.9.0: - added the ability for open4.spawn to take either an array of arguments or multiple arguments in order to specify the argv for the command run. for example
open4.spawn ['touch', 'difficult to "quote"'], :stdout=>STDOUT
same thing
open4.spawn 'touch', 'difficult to "quote"', :stdout=>STDOUT
thanks to jordan breeding for this suggestion
- added 'cwd'/:cwd keyword. usage is pretty obivous
open4.spawn 'pwd', 1=>STDOUT, :cwd=>'/tmp' #=> /tmp
this one also from jordan
0.8.0:
- fixed a critical bug whereby a process producing tons of stdout, but for
which the stdout was not handled, would cause the child process to
become blocked/hung writing to the pipe. eg, this command would cause a
hang
include Open4
spawn 'ruby -e" puts Array.new(65536){ 42 } "'
whereas this one would not
include Open4
spawn 'ruby -e" puts Array.new(65536){ 42 } "', :stdout=>StringIO.new
this version handles the former by spawning a 'null' thread which reads,
but does not process stdout/stderr. that way commands which generate
tons of output will never become blocked.
0.7.0: - merged functionality of exitstatus/status keywords:
include Open4
spawn 'ruby -e "exit 42"' # raises
spawn 'ruby -e "exit 42"', :status=>true # ok, returns status
spawn 'ruby -e "exit 42"', :status=>42 # raises if status != 42
spawn 'ruby -e "exit 42"', :status=>0,42 # raises if status != 0||42
- the 0.6.0 was broken on rubyforge... this release fixes that (somehow!?)
0.6.0: - added feature for exitstatus to be list of acceptable exit statuses
Open4.spawn 'ruby -e "exit 42"' # raises
Open4.spawn 'ruby -e "exit 42"', :exitstatus=>[0,42] # ok
- added :status switch, which will always simply return the status (no
error thrown for failure)
Open4.spawn 'ruby -e "exit 42"' # raises
status = Open4.spawn 'ruby -e "exit 42"', :status=>true # ok
note, however, that any SpawnError does in fact contain the failed
status so, even when they are thrown, error status can be retrieved:
include Open4
status =
begin
spawn 'ruby -e "exit 42"'
rescue SpawnError => e
warn{ e }
e.status
end
0.5.1: - fixes a critical but in ThreadEnsemble class that had a race condition that could cause thread deadlock. sorry bout that folks.
0.5.0: - on the suggestion of tim pease (thanks tim!), i added timeout features to open4. the command run may have an overall timeout and individual timeouts set for each of the io handles. for example
cmd = 'command_that_produce_out_at_one_second_intervals'
open4.spawn cmd, :stdout_timeout => 2
or
cmd = 'command_that_should_complete_in_about_one_minute'
open4.spawn cmd, :timeout => 60
or
cmd = 'consumes_input_at_one_line_per_second_rate'
input = %w( 42 forty-two 42.0 )
open4.spawn cmd, :stdin=>input, :stdin_timeout=>1
- added 'open4' alias so one can write
open4.spawn vs Open4.spawn
or even
open4(cmd) do |pid,i,o,e|
end
- added signal info to SpawnError
0.4.0: - improved error handling contributed by jordan breeding. - introduction of background/bg method
0.3.0 : - bug fix from jordan breeding. general clean up. added spawn method.
0.2.0 : - added exception marshaled from child -> parent when exec fails. thanks to jordan breeding for a patch (yay!) and paul brannan for this most excellent idea.
0.1.0 : - fixed docs to correctly show return value of popen4 (pid first not last). thanks Stefanie Tellex stefie10@alum.mit.edu for catching this. 0.0.0 : - initial version
AUTHOR
LICENSE
ruby's
FAQs
Unknown package
We found that open4 demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
Oracle seeks to dismiss fraud claims in the JavaScript trademark dispute, delaying the case and avoiding questions about its right to the name.
Security News
The Linux Foundation is warning open source developers that compliance with global sanctions is mandatory, highlighting legal risks and restrictions on contributions.
Security News
Maven Central now validates Sigstore signatures, making it easier for developers to verify the provenance of Java packages.